Paper ID #10008

Synchronized Robot: A PID Control Project with the LEGO Mindstorm

NXT
Dr. Chad Eric Davis, University of Oklahoma

Chad E. Davis received the B.S. degree in mechanical engineering, M.S. degree in electrical engineering,
and Ph.D. degree in engineering from the University of Oklahoma (OU), Norman, in 1994, 2000, and
2007, respectively. Since 2008, he has been a member of the Electrical and Computer Engineering (ECE)
faculty, University of Oklahoma. Prior to joining the OU-ECE faculty, he worked in industry at Uponor
(Tulsa, OK), McElroy Manufacturing (Tulsa, OK), Lucent (Oklahoma City, OK), Celestica (Oklahoma
City, OK), and Boeing (Midwest City, OK). His work experience ranges from electromechanical system
design to automation of manufacturing and test processes. His research at OU involves GPS ground-
based augmentation systems utilizing feedback control. Dr. Davis holds a dual discipline (electrical and
mechanical) professional engineering license in the state of Oklahoma. He currently serves as the faculty
adviser for Robotics Club, the Loyal Knights of Old Trusty, and Sooner Competitive Robotics at OU and
he serves as the recruitment and outreach coordinator for OU-ECE. He received the Provosts Outstanding
Academic Advising Award in 2010 and the Brandon H. Griffin Teaching Award in 2012.

Anh Mai, University of Oklahoma

Anh Mai received B.Sc. in Electrical Engineering in 2004 (Vietnam), and M.Sc. in Electrical Engineer-
ing in 2008 (University of Oklahoma). He is currently a Ph.D. candidate in Electrical Engineering at the
University of Oklahoma. His research interests include controls, robotics, artificial intelligence, dynam-
ics and controls of prosthetic devices. During his time at the University of Oklahoma, he involved in
development of embedded systems for applications in road construction and healthcare research.

American
c Society for Engineering Education, 2014
 Synchronized Robots:
A PID Control Project with the LEGO Mindstorm NXT
Abstract:

Typical courses in introduction to controls are usually heavy in theory and light in practical
application. In order to emphasize practical application in this type of course, such as the
implementation of a PID controller, there are problems with most of the commonly used options.
The first option is to provide the students with an approximate model of a system and have them
use simulation to implement a PID controller. The simulation approach is limited because there
is not an actual system to control and the model is likely oversimplified. A second option
requires the instructor to create a control system that all students use. This approach is
expensive, time consuming, and not all students have access to the system at the same time. A
third possibility requires students to build their own system and then control it. The problem
with this approach is that so much time is dedicated to mechanical and electrical design, while
important control theory concepts are left out or deemphasized. A solution to these problems is
providing the students with a LEGO Mindstorm NXT kit, which allows them to quickly design
their system so they can focus on the application of control theory. These kits are inexpensive
and can be used for other functions such as freshman orientation courses and outreach events
when not being used for controls.

In the Introduction to Control Systems Engineering course (ECE 4413) at the University of
Oklahoma (OU), students were given a LEGO Mindstorm NXT kit and some brief training in the
LabVIEW programming environment in order to complete a class project. The students
implemented PID control algorithms to produce synchronized robots that tracked each others
movement. Implementing PID control in LabVIEW without any built-in PID functions allows
students to discover how the control algorithm works in a visual way. With this method, the
likelihood of retaining the knowledge would appear to be much greater than modifying a built-in
function while doing simulations. This paper will show how this project increased the students
understanding of PID control. Pre-project and post-project quizzes are used to provide statistical
evidence of improvement. Students perceived value of the project is also explored.
Additionally, details of the project are provided so that other universities looking to emphasize
practical application in their introductory controls course can replicate it.

I. Introduction

The most practical topic in most controls courses is the implementation of a PID controller. For
instructors of control engineering classes that want to emphasize practical application, there are
many options for implementing PID control techniques into projects. The simplest way is to
create a project where PID control is explored through simulation 1-3. With simulations, learning
is somewhat limited because students are usually provided an oversimplified model and only
asked to make slight modifications to provided or existing code. Additionally, simulation
doesnt work well for a system that has a model that is unknown or difficult to attain, which is
usually what the students will face in industry.
Another option is for the instructor to create a complicated hands-on project and all of the
students control the same system at different times 4-8. There are many reasons why it is difficult
to create a control system that is dedicated for student projects. First, it is time consuming and
expensive to create and maintain them. Secondly, once a control system is created, the instructor
might have difficulty in managing the time for all students to work on it. The cost would be
highly increased if the instructor decides to create numerous identical control systems. The third
issue is that current students can get project information from previous students instead of
learning it for themselves since the system will likely not change from semester to semester.

Another possibility is the students build the system from scratch and then control it9-11. The
problem with this approach is that so much time is dedicated to mechanical and electrical design,
while important control theory concepts are left out or deemphasized. The idea presented in this
paper is to use the final option of having the students build their own system, but provide them a
kit so they can focus more on implementing the PID controller.

II. ECE 4413 Course Information

ECE 4413 is a senior level elective course that is titled: Introduction to Control Systems
Engineering. This introductory course is a typical first course in controls that covers topics such
as: modeling linear systems using the Laplace and State-Space methods, time response
characteristics, stability, feedback, system performance, frequency response, block diagram
simplification, and controller design with emphasis on the PID algorithm. The pre-requisite for
ECE 4413 is a junior-level course titled ECE 3793: Signals and Systems. In all of the
textbooks the authors reviewed and the control system courses they have been involved with in
the past, PID control is a topic that is always covered toward the end of the semester. The first
problem with this late coverage is that it usually makes it difficult to have a PID control project.
Secondly, since PID control is the most practical aspect of this theoretical course, students can
easily find themselves losing interest because they dont see how the information they are
learning applies to the real world until the PID control section is covered. In order to prevent
these negative outcomes, the material in ECE 4413 in the Fall 2013 semester was taught in a way
that allowed PID control to be covered in the middle of the semester instead of the end. Topics
(such as state space modeling, sensitivity, disturbances, and advanced block diagram reduction
techniques) that are not required to be covered in order to understand PID control were moved to
the end of the semester, while feedback control systems and root locus design techniques were
covered earlier. In order for models to be created for electrical and mechanical systems at the
beginning of the semester, the students were taught techniques that do not involve block diagram
reduction, such as differential equations and matrix techniques utilizing the Laplace transform.

Another objective in this course is to provide the students an opportunity to use Matlab and
LabVIEW software in a more applied way than previous courses. Matlab is used throughout the
semester to provide the tools needed to solve the homework assignments and a method to check
their work when asked to solve problems without a computer. LabVIEW was used primarily in
team projects that will be discussed in more detail later in this paper. When forming a team it is
important to keep the project groups small so that every student has the opportunity to work on
the LabVIEW software. However, having multiple members provides valuable experience in
working on team projects. One solution to this dilemma is setting up teams like a FIRST
Robotics competition (FRC) where there are individual teams and multiple individual teams
make up an alliance. In ECE 4413 each individual team was comprised of two students and two
individual teams worked together on an alliance. Each team would create a robot that would
work with its alliance teams robot. The Synchronized Robot project described in this paper
works perfectly for this team structure.

III. LEGO Mindstorm NXT Platform Description

For the idea discussed in this paper to be implemented by others, the kit that students are given to
complete their robotics project will first be described. All parts are put in a small lightweight
toolbox that the students can easily carry around with them. The LEGO Mindstorm NXT kit
(Part # 8547) includes 1 32-bit microprocessor LEGO NXT brick, a battery with charger, a USB
cable to connect the NXT to the computer, 3 interactive servo motors with internal encoders, 2
light sensors, 2 touch sensors, 1 sound sensor, and 1 ultrasonic sensor. Many additional sensors
are also available from LEGO. The LEGO Mindstorm NXT kit also includes an adequate supply
of LEGO pieces. The light sensors can be used in reflective mode with an internal LED source
or in ambient mode, where a flashlight or other form of external lighting can be used to control
the robot. The students are then required to install LabVIEW and the LEGO Mindstorm NXT
Module on their laptops, both of which can be downloaded from NI.com. The LEGO Mindstorm
NXT has proven to be a versatile platform as they have been used at our university in numerous
ways and for varying student levels. The same platform has also been used for hands-on
activities for high school outreach events and freshman orientation projects. Using the LEGO
Mindstorm NXT platform with LabVIEW has allowed it to be used for engineering courses,
Robotics Club events, and even LabVIEW training seminars for graduate students.

IV. Phase 1 Robot Project LabVIEW State Machine Architecture

One difficulty in implementing this project is that many of the students in the course are not
proficient in LabVIEW. In fact, roughly one-third of the class had no LabVIEW experience at
all. With experience doing a similar project to the Phase 1 Robot Project 12 in a different course,
it was determined that some LabVIEW training would be needed. Therefore, LabVIEW tutorials
were provided for the students in order for them to reach a modest level of proficiency. The
LabVIEW training started at the beginning of the course and was done in parallel with control
theory topics that were presented in lecture. Once the students completed this training and
finished a LabVIEW state-machine homework assignment, the LEGO Mindstorm NXT kits were
handed out. One lecture was dedicated to getting the teams and alliances set, handing out the
Mindstorm kits, and showing the students how to program their LEGO Mindstorm using the
LabVIEW NXT module. Next, they were given their first challenge: Make the two alliance
robots travel back and forth in a synchronized fashion using a state-machine architecture. The
project will be repeated in Phase 2 using PID control. Requiring the students to first do the
synchronization with state-machines allows them to get comfortable programming in LabVIEW
and see how much better results they can achieve when using PID control techniques later in
Phase 2. Furthermore, the Phase 1 Project was only worth ~ 6% of their final grade, but Phase 2
was worth 20% of their final grade.
 A. Setting up the robots for the competition

The following figure shows how each of the alliances robots were set up in the competition.

Ultrasonic Sensor

Light Reflectivity Sensor Follower Robot

Leader Robot L-Shaped Wall

Figure 1: Synchronized Robot Competition Overview (Used for both Phase 1 and Phase 2)

The alliances work together to create an L-Shaped wall that can be used in combination with the
follower robots ultrasonic sensor to determine the distance between the robots. The goal is for
the follower robot to maintain a fixed distance between it and the leader robot. The desired
distance needs to be measured so that the followers light sensor is over the strip of silver tape.
The amount of time the light sensor is over the silver tape is calculated by the follower scoring
LabVIEW program (VI), which is shown Figure 2 below. The students build their state machine
program inside this VI template.

Figure 2: Phase 1 Follower Robot Template and Scoring VI

The leader robot is also programmed inside a template, as shown in the Figure 3. This template
also includes the scoring measurement of the total radians of movement of the leader robot. The
final score is calculated by multiplying the radians of movement by the percentage of the time
the follower was over the silver tape. This final scoring metric determines the amount of
movement where the robots were synchronized (within the desired range of distance of each
other).

Figure 3: Phase 1 Leader Robot Template and Scoring VI

B. Competition Rules

The robots have 30 seconds to build up their score by traveling back and forth across a round
black panel that has an approximate diameter of 4 with 2 white ring around the outside edge.
The panel is raised so that if either robot crosses the white outer ring they will fall off and not be
able to increase their score further. Each alliance of two teams performs two runs. Each team of
an alliance acts as a leader in the first run then switches its role to a follower in the second run.
These two scores are averaged and bonus points awarded to the alliance that has the highest
average score.

V. Phase 2 Robot Project LabVIEW PID Control

The Phase 2 Project was much more complicated than the Phase 1 Project since a PID control
algorithm was required to be programmed into LabVIEW and there were two separate challenges
in the project (see Part 1 and Part 2 below). Depending on the complexity of the LabVIEW
programs they created, the sampling rate and accuracy of digitization might be varied from group
to group. Therefore, we do not focus on specifying the detailed characteristics of digital control
implementation in this paper.

The following two parts of the project were combined into one final score.
 Part 1 For this part, the students did the exact same thing as the Phase 1 Project except
they were required to use PID control in the follower instead of a state-machine. New
LabVIEW template files were created in a similar fashion to those shown in Figure 2 and
Figure 3. The main difference is that there are two sequential while loops for parts 1 and
2. An example of one of the teams LabVIEW code for the follower robot for part 1 is
shown in Figure 4.

Figure 4: Example of Student Phase 2 Follower Robot Part 1 LabVIEW Code

Part 2 After the 30 seconds expires from Part 1, a new while loop will execute in the
template. The follower is required to go to the edge and travel all the way around the
panel by tracing the white ring without falling off. During this time, at least a part of the
robot must stay within one inch of the white ring. After the follower robot finishes one
trip around the panel, the orange button on the NXT brick is pressed and the scores are
read. The leader robot tries to stay in the panel, but not run into the follower. This
requires additional synchronization between the robots. If the robots got out of sync they
could run into each other and the follower would not be able to successfully complete the
revolution around the panel, which would result in a major score deduction (scale factor
of 0.1). Some teams used a state-machine architecture for Part 2 and some used PID
control. Figure 5 shows the follower LabVIEW code for one of the teams that used PID
control for Part 2.

The scoring for the Phase 2 Project is described below. The alliances that had the best Part 1
score and the individual teams that had the best Final Score were awarded bonus points.
The Part 1 scoring is the similar to Phase 1 Project: (Silvertime/30)*Total_Movement.
The Part 2 scoring is the time it takes the follower robot to find the white ring and finish one
trip around the panel. This time is used to adjust the Part 1 score based on the team ranking.
 o FINAL SCORE = Part1_Score * Scale_Factor
o If Part 2 is not successfully completed then the scale factor is 0.1.
o If Part 2 is successfully completed then the scale factor is determine by the team
ranking as compared to other students in the class.

Figure 5: Example of Student Phase 2 Follower Robot Part 2 LabVIEW Code

VI. Procedure for Project Preparation and Performance Assessment

In this section, the comprehensive procedure used in the class project is described. With this
procedure (Figure 6), the instructor was able to (1) prepare necessary background and skills for
students prior to the project and (2) evaluate the competency and confidence of students upon
project completion.
 Instructor prepares backgrounds and
skills for students

Students take Bonus Quiz 1 on the PID

control algorithm with LabVIEW

Students implement PID control on the

NXT for the project

Students take Bonus Quiz 2 on the PID

control algorithm with LabVIEW and ~
6 weeks later take the final exam
Figure 6: Step-by-Step Procedure of the Project

A. Background and Skill Preparation for the Project

In order to successfully implement the PID control algorithm using LabVIEW on the LEGO
Mindstorm NXT and use the robots for the class competition, students were required to gain
some backgrounds/skills through the following tasks (Figure 7).
Task 1: Complete the project using state machine in LabVIEW (Phase 1 Project). Upon
completion, students were expected to increase their competency in LabVIEW, learn how
to setup the project as shown in Figure 1, and how the scoring schemes were executed in
the LabVIEW template program.
Task 2: Simulate the PID control algorithm. Students were required to complete the take-
home midterm problem where they had to simulate PID control algorithm in Matlab.
Upon completion, students were expected to understand the principles of the PID control
algorithm as well as practical tuning methods for the proportion gain Kp, integration gain
Ki, and derivative gain Kd such as Ziegler-Nichols.
Task 3: Implement PID control in LabVIEW. Students were instructed to implement a
PID control algorithm in LabVIEW. By doing a homework assignment in LabVIEW,
students practiced to obtain major components of the PID control algorithm such as error
between the output feedback and the reference, temporal difference for the derivative
part, or accumulation of error for the integral part of the controller. The students also
experimented with anti-windup and gain scheduling schemes.
 At this point, students
should be ready to
Task 1
implement PID control
Complete the
on the NXT Robots
project using state
using LabVIEW
machine in
LabVIEW

Task 3
Task 2 Implement PID
Simulate PID control in
control algorithm LabVIEW

Figure 7: Background/skill Preparation for the Project

B. Bonus Quizzes
In order to evaluate the improvement of students scores as well as their confidence in
implementation of the PID control algorithm in a practical device such as the NXT robot, bonus
quizzes with similar formats were given before and after the Phase 2 Project.
Students were given 4 block diagrams which implemented the PID control algorithm in
LabVIEW. Only one block diagram implemented the PID control correctly (see Figure 8
and Figure 9).
In order to test for correct PID control structure, students were asked to select the block
diagram that implemented the PID control algorithm correctly.
In order to test for correct PID control parameters, students were asked to find the control
gains Kp, Ki, and Kd from the block diagram they selected.
Student selected how they wanted to be graded in the bonus quizzes by answering the
final question: Do you want to go for 10 points with the potential of drop down to 1
point if any of your answers are wrong? (Selecting No indicates you are opting for 5
points) Circle Yes or No
Students grades were determined using the following rules
o 10 points Answer Yes on the final question and write down correct control gain
Kp, Ki, Kd for the correct block diagram of the PID control algorithm
o 5 points Answer No on the final question
o 1 point Answer Yes on the final question and either select incorrect block
diagram or enter incorrect control gains.
It is important to note that students were not allowed to see neither the solutions nor the
results of the first bonus quiz. In addition, after the project completion, the second bonus
quiz was given with a similar format, but the LabVIEW block diagrams were wired
differently. Both of the quizzes were given without advance notice, therefore students
were not able to study for the quizzes and just answered with knowledge and skill that
they possessed at that time.

Figure 8: Correct PID Block Diagram

(a)

(b)
 (c)

Figure 9: Incorrect PID Block Diagrams - (a) Incorrect P (Connect to de instead of e); (b)
Incorrect I (Integration not Accumulated); (c) Incorrect D (Connect to t instead of dt)

C. Final Exam LabVIEW Question

On the final exam, in addition to the standard questions, students were given a LabVIEW block
diagram that implemented the PID control algorithm incorrectly. They were asked to correct the
diagram (test for PID control structure) and find the control gains (test for PID control
parameters). The final exam was approximately 6 weeks after the Bonus Quiz 2 so it showed
whether the students retained the knowledge. It is important to note that the bonus quizzes were
not returned to the students until after the final exam so the students had to rely on the
knowledge they gained from the Phase 2 Project instead of studying the answers to the questions.

VII. Results

A. Selection Criteria
Out of total 30 students in the class, there were 21 students who met all of the following
selection criteria and gave written consent to participate in this study.
1. Had not implemented PID control algorithm in the LEGO Mindstorm NXT before the
Bonus Quiz 1: 3 out of 30 students did not meet this selection criterion.
2. Showed up in class and answered both bonus pre-project and post-project quizzes: 5 out
of 30 students did not meet this selection criterion.
3. Showed up for the final exam: 1 out of 30 students did not meet this selection criterion.

B. Evaluation Methods
The first evaluation metric were the grades from two bonus quizzes and one LabVIEW question
on the final exam.
1. Students grades and their selection on the final question of the bonus quizzes are
collected and organized in
2. Table II.
3. Students who answered both PID structure questions and PID parameter questions
correctly were considered to achieve 100% correct in
4. Table II.

In addition to the bonus quizzes and final exam question, the students self-perceived value of
the synchronized robot project was assessed in a survey shown in Table I. The survey is on a 5
point Likert scale where 5 is the most favorable response and 1 is the least favorable response.
For question Q1, Q2, Q4, and Q6, students responses were associated to three categories:
negative (responses 1 and 2), neutral (response 3), and positive (responses 4 and 5). These
categories are used for some of the data analysis described later.

Table I - Survey Questions and Reponses

Question Average
Response
(Q1) The NXT Synchronized Robot Project helped me understand PID control more than 4.0
the homework assignment where I did Matlab simulations using PID control.
(Q2) If asked to implement a PID control algorithm in the future I am confident I could do it. 3.9
(Q3) The NXT Synchronized Robot Project was very difficult. 2.7
(Q4) Working with my group and alliance greatly improved my ability to work on a team. 3.6
(Q5) I would rate my competency in LabVIEW prior to this class the following number. 2.5
(Q6) This project greatly improved my competency in LabVIEW. 3.3

C. Data Analysis
Data presented herein are from the 21 students who met the selection criteria as described
previously. Shaded numbers in
Table II correspond to Final Question Yes responses.
Table II Results from Bonus Quizzes and Final Exam
Score 100% correct
Student
BQ1 BQ2 BQ1 BQ2 FE
S1 5 5
S2 1 10
S3 5 5
S4 5 5
S5 5 10
S6 5 5
S7 5 10
S8 5 5
S9 5 5
S10 5 10
S11 5 10
S12 10 10
S13 1 1
S14 5 10
S15 5 5
S16 5 5
S17 5 1
S18 5 1
S19 5 5
S20 5 10
S21 10 10
Note: BQ1, BQ2 = Bonus Quiz 1, 2; FE = Final Exam
100% correct = (correct PID structure + correct PID parameter)
Statistical Analysis: In order to answer the question of whether or not the project helped
students improve their score between the Bonus Quiz 1 (pre-project) and the Bonus Quiz 2 (post-
project), statistical analysis was carried out to test the hypothesis that students scores before
the project are less than their scores after the project. One-tailed, paired t-test ( = 0.05)
between the scores of pre-project and post-project quizzes supported the hypothesis with the
probability of accepting the null hypothesis was only 0.026716 (less than ). This concludes that
the project helped students improve their scores in the bonus quizzes.

In addition, the percentage of students who answered both questions on PID control structure and
PID control parameters correctly (referred as 100% correct in
Table II) also increased from ~9.5% (2 out of 21) before the project to ~42.9% (9 out of 21) after
the project and ~57.1% (12 out of 21) in the similar LabVIEW question in the final exam.

Results in
Table II also answered the question about the students confidence in accepting the challenge by
selecting Yes on the final question of the bonus quizzes. Percentage of students who chose to
accept the challenge increased from ~19% (4 out of 21) before the project to ~57.1% (12 out of
21) after the project. Among the students who accepted the challenge, the percentage of getting
100% correct increased from 50% (2 out of 4) before the project to 75% (9 out of 12) after the
project. Students responses to survey questions are shown in Table I and Figure 10. Numbers
of responses to questions Q4 and Q6 did not add up to 21 since one student did not specify his
answers.

Figure 10: Students Responses to Survey Questions

It is very promising to observe that by completing a not-very-difficult project (in Q3, only
~23.8% of students, or 5 out of 21, agreed that the project was very difficult) such as the
Synchronized Robot Project, students reported improvement in many aspects.
In Q1, ~66.7% of students (14 out of 21) said that the project helped them understand
PID control more than the homework assignment where they simulated PID control
algorithm using Matlab. Only ~4.8% of students (1 out of 21) disagreed.
 In Q2, ~66.7% of students (14 out of 21) felt confident to implement the PID control
algorithm in the future. None of the students disagreed.
In Q4, ~52.4% of students (11 out of 21) said that working with their group and alliance
greatly improved their ability to work on a team. Only ~4.8% of students (1 out of 21)
disagreed.
According to responses to Q5, prior to the class, only ~33.3% of students (7 out of 21)
had a high level of competency in LabVIEW compared to ~47.6% of students (10 out of
21) had a low level of competency. After the project, ~52.4% of students (11 out of 21)
said that their competency in LabVIEW had been greatly improved by doing the NXT
Synchronized Robot Project (response to Q6). It is noted that all of ~23.8% of students
(5 out of 21) who disagreed with Q6 already had high level of competency in LabVIEW
prior to this class.

VIII. Conclusion

This paper shows that using the LEGO Mindstorm NXT offers advantages over using other
alternatives for a PID control project. By providing the students with an interesting and
challenging robot competition the students were able to learn PID control concepts. The pre-
project and post-project quizzes and responses to the final exam question show that the project
increased the students knowledge in how to correctly implement a PID control algorithm. The
students also self-reported in the survey that the project was beneficial to them in the following
ways: 1) Helped them understand PID Control, 2) Gave them confidence they could implement
PID control in the future, 3) Improved their ability to work on a team, and 4) Improved their
competency in LabVIEW. Additional evidence that the project was effective came from the
course evaluations that 18 students completed. To the statement In general, the instructor
taught this course effectively an average response of 4.78 out of 5 was reported, which was 15%
higher than the average response to this question from all College of Engineering course
evaluations at OU. To the statement The way in which the course was taught helped me
develop analysis and design abilities an average response of 4.61 out of 5 was reported, which
was 18% higher than the average response to this question from all College of Engineering
course evaluations at OU. One student commented: Integration of LabVIEW programming into
the course gave another angle on the way control systems can be implemented. The qualitative
and quantitative results from this project suggest that it is worthy of consideration for others
looking to emphasize practical application in their introductory controls course.

IX. References
1. S. Ramasamy, H. V. Pradhan, P. Ramanathan, P. Arulmozhivarman, and R. Tatavarti, "A novel and pedagogical
approach to teach PID controller with LabVIEW signal express," 2012 IEEE International Conference on
Engineering Education: Innovative Practices and Future Trends (AICERA), pp.1-8, 2012.
2. J. L. Newcomer, "A design project based approach to teaching automatic control theory to mechanical
engineers," 28th Annual Frontiers in Education Conference, 1998. FIE '98., vol.3, pp.1242-1245, 1998.
3. A. O'Dwyer, " An approach to teaching PID controller tuning," Proceedings of the 7th IFAC Symposium on
Advances in Control Education, 2006.
4. J. A. Rehg, "Design of a low cost trainer for flow control," 28th Annual Frontiers in Education Conference,
1998. FIE '98., vol.3, pp.1063-1067, 1998.
5. R. P. Ramachandran, R. Ordonez, S. Farrell, Z. O. Gephardt, and H. Zhang, "Multidisciplinary control
experiments based on the proportional-integral-derivative (PID) concept," Proceedings of the 2001 American
Society for Engineering Education Annual Conference & Exposition, 2001.
6. N. K. Kim, "Process control laboratory experiments using LabVIEW," Proceedings of the 2001 American
Society for Engineering Education Annual Conference & Exposition, 2001.
7. D. Um and V. Sriraman, "Teaching basic control systems theory using robots," Proceedings of the 2004
American Society for Engineering Education Annual Conference & Exposition, 2004.
8. S. Daniels, D. Harding, and M. Collura, "Introducing feedback control to first year engineering students using
LabVIEW," Proceedings of the 2005 American Society for Engineering Education Annual Conference &
Exposition, 2005.
9. C. Chen and J. Derr, "Radio-controlled robot lawnmower," Proceedings of the 2004 American Society for
Engineering Education Annual Conference & Exposition, 2004.
10. R. L. Mueller, "The inverted pendulum problem as a senior design project," Proceedings of the 2004 American
Society for Engineering Education Annual Conference & Exposition, 2004.
11. G. E. Meyer, "An improved instrumentation and controls course for agricultural and biological engineering,"
Proceedings of the 2009 Midwest Section Conference of the American Society of Engineering Education, 2009.
12. C. E. Davis and P. S. Pulat, "Redesigning the circuits for non-majors course with the addition of a robotics
project," Proceedings of the 2013 American Society for Engineering Education Annual Conference &
Exposition, 2013.